question: RAG by definition offloads the retrieval to a vector similarity search via embeddings db (faiss, knn et al). what is the preferred way to feed docume

question: RAG by definition offloads the retrieval to a vector similarity search via embeddings db (faiss, knn et al). what is the preferred way to feed documents / knowledge into a model so that the primary retrieval is done by the llm, and perhaps use vector db only for information enhancement (a la onebox)?